Polytechnic University of Bucharest

Due to changes in global climate, coastal erosion is a common phenomenon, leading to flooding, habitat loss, property damage, and economic impacts. Large-scale automatic monitoring of coastlines is feasible by processing satellite images with Artificial Intelligence (AI) models; however, it is often constrained by scarce labels and model opacity. We present an explainable, fully unsupervised framework that first converts multi-year Sentinel-2 image series into pixel-wise land–water maps with attached model certainty and then discovers change patterns in the temporal structures of these land-water masks over a specified study period. Throughout the pipeline, we used the Normalized Difference Water Index (NDWI), computed from Sentinel-2 Green (B3) and NIR (B8), as our primary spectral indicator of surface water. We leverage two probabilistic models independently to select the best: (i) a Latent Dirichlet Allocation (LDA) model on a bag-of-visual-words to discover interpretable topics (e.g., foam, water, vegetation) that provide semantic segmentation with transparency, and (ii) a Gaussian Mixture Model (GMM) in pixel feature space to produce per-pixel probabilistic clustering and segmentation. Both models create multiple binary land-water segmentation maps from static images and then a single time-series evolution map from these multi-year land-water segmentation maps. The shoreline was automatically extracted as the boundary land pixels in binary maps with a threshold-based certainty score. We summarize changes with three methods and a score to define the amount of changes: the scene-level water fraction <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p_w</i>, shoreline-normal displacement (SND) along fixed transects, and the low-confidence proportion <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">U</i>τ that localizes ambiguity to physically dynamic interfaces. A Hamming score was computed to provide a quick and quantitative overview of the time-series evolution over a 7-year period. The pipeline delivers clear coastlines, interpretable segmentation maps, and uncertainty visualizations to help domain experts in auditing decisions. The result is a label-free, scalable, and explainable workflow in which every decision is accompanied by a confidence score, supporting reliable coastal monitoring and downstream scientific and application-oriented use.

This study presents a low-cost irrigation prototype developed on an Arduino platform, integrating soil moisture sensors and a fuzzy logic controller. The system was tested under contrasting weather scenarios to evaluate its capacity to stabilize soil water content. Results showed clear differences between cloudy and sunny conditions, with the latter leading to accelerated drying and reduced infiltration. Soil sensor data confirmed distinct spatial patterns of moisture distribution, while statistical analysis revealed a strong negative correlation between soil moisture and irrigation volume. The fuzzy algorithm responded proportionally to water deficits, applying larger volumes during recovery phases and frequent low-volume cycles for stabilization. These findings demonstrate the potential of fuzzy logic-based irrigation to enhance water efficiency and maintain soil moisture balance under variable environmental stresses.

Deep neural networks currently dominate many fields of the artificial intelligence landscape, achieving state-of-the-art results on numerous tasks while remaining hard to understand and exhibiting surprising weaknesses. An active area of research focuses on adversarial attacks, which aim to generate inputs that uncover these weaknesses. However, this proves challenging, especially in the black-box scenario where model details are inaccessible. This paper explores in detail the impact of such adversarial algorithms on ResNet-18, DenseNet-121, Swin Transformer V2, and Vision Transformer network architectures. Leveraging the Tiny ImageNet, Caltech-256, and Food-101 datasets, we benchmark two novel black-box iterative adversarial algorithms based on affine transformations and genetic algorithms: 1) Affine Transformation Attack (ATA), an iterative algorithm maximizing our attack score function using random affine transformations, and 2) Affine Genetic Attack (AGA), a genetic algorithm that involves random noise and affine transformations. We evaluate the performance of the models in the algorithm parameter variation, data augmentation, and global and targeted attack configurations. We also compare our algorithms with two black-box adversarial algorithms, Pixle and Square Attack. Our experiments yield better results on the image classification task than similar methods in the literature, achieving an accuracy improvement of up to 8.82%. We provide noteworthy insights into successful adversarial defenses and attacks at both global and targeted levels, and demonstrate adversarial robustness through algorithm parameter variation.

Recent advances in sensors for medical applications enhance data acquisition for medical systems. The authors propose a Digital Twin (DT) architecture for glaucoma diagnosis and monitoring that integrates Contact Lens Sensors for real-time monitoring of glaucoma progression. The proposed framework is discussed in regard to ethical aspects associated with the use of Digital Twin technologies in medicine. The use of emerging technologies in medicine is associated with significant ethical challenges. Starting from this premise, an analysis is conducted on data traceability, and several questions are raised regarding the responsible integration of predictive technologies into diagnostic and treatment decision-making processes. However, in order to combine the respect for fundamental ethical principles with the maximization of benefits, this paper proposes a guideline that integrates a set of deontological rules and utilitarianism rules.

The very fast digitalization of electrical systems has brought remarkable opportunities and significant risks. Infrastructures that were once largely isolated are now significantly interconnected, relying on Supervisory Control and Data Acquisition (SCADA) platforms, Industrial Internet of Things (IIoT) devices, and cloud-based services. Considering this transformation has improved efficiency and flexibility, it has also broadened the attack surface available to malicious actors. Highprofile incidents over the past decade, such as the Stuxnet worm in 2010, the cyberattacks on Ukraine's power grid in 2015-2016, and the SolarWinds supply chain compromise in 2020, have exposed the vulnerability of power systems to sophisticated threats. This paper explores resilience and vulnerability as two complementary aspects of protecting electrical infrastructures. Vulnerabilities are examined from many perspectives, such as technical, organizational, and supply-chain, while resilience is understood as the ability to anticipate, withstand, and recover from many disruptive events. More particular attention is given to emerging approaches, including the integration of artificial intelligence (AI) with Security Information and Event Management (SIEM) tools, the use of microgrids for redundancy, and the role of cross-border regulatory frameworks. The discussion is highlighting the need for a holistic approach that combines technology, governance, and international cooperation. Such approach is essential for building adaptable, flexible infrastructures that are capable of maintaining energy security in an increasingly complex and uncertain threat landscape. The major importance of this work is that it brings together cybersecurity and engineering perspectives, showing how real incidents are translated into measurable impacts on power systems and suggesting practical ways such as AI based monitoring and digital twin models, to empower their resilience.

This study explores an innovative delivery strategy for the management of skin conditions: lipid nanosystems incorporated into a gel matrix. Echinacea purpurea extract, known for its antibacterial, antioxidant, and wound-healing properties, was encapsulated into lipid-based nanosystems and subsequently incorporated into Carbopol-based gel. The extract, rich in chicoric and caftaric acids, exhibited strong antioxidant activity (IC50 = 56.9 µg/mL). The resulting nanosystems showed nanometric size (about 200 nm), high entrapment efficiency (63.10–75.15%), and excellent short-term stability. Superior biocompatibility of the nanosystems, compared to the free extract, was demonstrated using an MTS assay on L-929 fibroblasts. Moreover, the cytoprotective potential of the lipid carriers was evident, as pre-treatment significantly increased cell viability under H2O2-induced oxidative stress. These findings suggest that lipid-based encapsulation enhances the therapeutic profile of E. purpurea. The optimal lipid formulation was incorporated into a Carbopol-based gel, which demonstrated an appropriate pH (5.15 ± 0.75), favorable textural properties, sustained polyphenol release, and overall good stability. This research highlights the potential of plant-derived bioactives in the development of dermatocosmetic products, aligning with current trends in eco-conscious and sustainable skincare.

The primary objective of the present study is to evaluate the effect of conglomerate microorganisms on nitrification in activated sludge. The present study compares this process with activated-sludge technology to explore the variables that influence the complex biochemical processes taking place in bioreactors. The research under consideration involves monitoring the effectiveness of optimizing the wastewater treatment process using kinetic modeling for the nitrification and denitrification processes. The system is designed to simulate various operating scenarios and adjust process parameters in real time. The nitrification rate demonstrates a 99.03% performance, while the denitrification rate ranges from 19.08% to 91.01%. A substantial correlation has been demonstrated between this variable and the temperature of the treated wastewater. This provides the possibility of accurately assessing the ammonium oxidation potential. Furthermore, kinetic equations facilitate the estimation of parameters that are not typically measured, yet are essential for optimizing operational parameters (e.g., dissolved oxygen levels in the aeration tank, sludge dosage, and influent flow rate). This estimation is crucial for enhancing the effectiveness of the process and attaining the desired or anticipated outcomes. This validation underscores the efficacy of the technology, thereby establishing a foundational framework for subsequent research endeavors. These research efforts are directed towards providing decision-makers and stakeholders with actionable insights. The validation underscores the significance of optimized practices in the context of water resource protection. Moreover, it signifies a substantial advancement in the instrumentation of wastewater treatment plants.

Abstract Background Cobalt-chromium (Co-Cr) alloys are extensively used in medical implants for their strength and corrosion resistance. However, improving their biocompatibility is essential to ensure better integration and performance in biological environments. This study investigates the effect of surface roughness and hydroxyapatite (HA) coatings on enhancing the biocompatibility and antimicrobial properties of Co-Cr alloys, aiming to create a durable, bioactive surface suitable for long-term implant use. Method Co-Cr alloy samples were prepared with varying levels of surface roughness achieved through grit-blasting with Al₂O₃ particles of different sizes (29 µm, 45 µm, 110 µm, and 250 µm). HA coatings were applied using a sol-gel spin-coating method to ensure even coverage and adhesion. The coated samples were analyzed using Scanning Electron Microscopy (SEM) and X-ray Diffraction (XRD) for morphological and structural assessment. Biocompatibility and antimicrobial properties were evaluated through immersion in Simulated Body Fluid (SBF) and testing against Pseudomonas aeruginosa biofilm formation. Results HA-coated samples exhibited high homogeneity and adhesion, particularly on rougher surfaces (110 µm and 250 µm). SEM and XRD analyses confirmed enhanced stability and crystalline structure on roughened surfaces. Biological tests demonstrated that the HA-coated samples with higher roughness significantly inhibited Pseudomonas aeruginosa biofilm formation, indicating improved antimicrobial effectiveness. The samples also showed sustained stability in SBF, supporting their potential for long-term application in implants. Conclusion Surface roughness positively impacts the adhesion and effectiveness of HA coatings on Co-Cr alloys, enhancing biocompatibility and antimicrobial performance. HA-coated Co-Cr alloys with optimized roughness provide a promising solution for medical implants, offering both durability and infection resistance, crucial for long-term clinical applications.

In recent years, significant attention has been directed toward the development of coating materials capable of tailoring surface properties for various functional applications. Transition metal nitrides, in particular, have garnered interest due to their superior physical and chemical properties, including high hardness, excellent wear resistance, and strong corrosion resistance. In this study, a fabrication process for CrN-based thin films was developed by combining reactive direct current magnetron sputtering (dcMS) with post-deposition annealing in air. CrN coatings were deposited by reactive dcMS using different argon-nitrogen (Ar:N2) gas ratios (4:1, 3:1, 2:1, and 1:1), followed by annealing at 550 °C for 1.5 h in ambient air. XRD and EDS analysis revealed that this treatment results in the formation of a composite phase comprising CrN and Cr2O3. The resulting coating exhibited favorable mechanical and tribological properties, including a maximum hardness of 12 GPa, a low wear coefficient of 0.254 and a specific wear rate of 7.05 × 10-6 mm3/N·m, making it a strong candidate for advanced protective coating applications.